Hewlett Packard Enterprise's Backlog of Nearly $6 Billion Is Fueled by a New Wave of AI Spending

Hewlett Packard Enterprise (NYSE: HPE) has gone from a legacy hardware vendor to an artificial intelligence (AI) infrastructure player in a matter of months. The stock is up 81% year to date, and management recently raised full-year earnings guidance by over 40% after the company blew past expectations in the second quarter.

While the first wave of AI infrastructure spending was dominated by hyperscalers building massive cloud data centers, the second phase is being driven by enterprises building their own on-premises AI capabilities. Running AI workloads with a variety of models on your own hardware is cheaper, and allows companies to protect their intellectual property, data, and competitive advantages.

HPE's timely acquisition of Juniper Networks last year positioned it to benefit from this spending. Businesses are drawn to Hewlett Packard Enterprise's integrated approach, which combines servers, storage, and high-performance networking gear, allowing its customers to build AI factories they control.

Image source: Getty Images.

Why networking drives deal size

Running AI requires graphics processing unit (GPU) clusters and networking hardware that communicate without delays. If the network lags, expensive GPUs sit idle.

After adding Juniper's capabilities, HPE can now offer a complete, integrated stack of compute, networking, storage, and private cloud software. Management noted on its second-quarter earnings call that demand for Juniper's solutions is now pulling through larger deals for servers and storage. Networking revenue reached $2.7 billion in Q2, with segment operating margins of 21.6%, accounting for over 40% of the company's total operating income.

As its networking solutions open the door for larger infrastructure sales, HPE is positioned to improve its profit margins as it captures a growing share of enterprise budgets. Competition from larger rivals such as Cisco and Arista Networks will be stiff, but broad-based demand should keep HPE busy.

Taking traditional servers along for the ride

Traditional server orders tripled in the second quarter, as companies aim to build out inference and agentic AI capabilities. HPE exited the quarter with a record $5.9 billion backlog, as demand for its AI systems and traditional servers is growing faster than it can ship them.

The jump in orders supports HPE's strategy to become the preferred provider of on-premises AI servers, but the company will need to work through industrywide supply shortages of components such as memory to convert its growing backlog into revenue.

For investors, the stock is not as attractive a buy as it was just a few months ago. That said, trading at roughly 13 times this year's earnings estimates, it's still a solid investment on a theme that's still in its early stages.

Bryan White has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Arista Networks, Cisco Systems, and Hewlett Packard Enterprise. The Motley Fool has a disclosure policy.